Results 401 comments of roryqi

Do Flink have similar trace system? Spark don't have similar trace system. We ever implement this feature, but we don't choose to merge it. We just use that feature to...

Could you add some performance results? Should we only modify the critical code?

> Any progress? Do i need to do what before merging @jerqi I think we need some performance tests to prove the effectiveness of change.

> I think we could directly follow the Spark change. Besides, the performance test looks hard. @jerqi I prefer not merging this pr without test. Maybe we can solve this...

> At present, there is only disk i/o in the health detection of the machine. Whether to add the index of memory utilization? What do u think of that? @jerqi...

Spark has exchange reuse. It may be very complex for fine-grained shuffle manager fallback.

When we reuse the Shuffle-ID-1, the rss don't have enough server nodes, what will it happen?